Open AkiraTOSEI opened 3 years ago
Apply noise and style vector optimization to the latent space of StyleGAN, we were able to achieve a more detailed image conversion. Two images can be stitched together naturally, even when they are not face images.
https://arxiv.org/abs/1911.11544
TL;DR
Apply noise and style vector optimization to the latent space of StyleGAN, we were able to achieve a more detailed image conversion. Two images can be stitched together naturally, even when they are not face images.
Why it matters:
Paper URL
https://arxiv.org/abs/1911.11544
Submission Dates(yyyy/mm/dd)
Authors and institutions
Methods
Results
Comments